Projects / selected work
Research, developer tools, educational software, and experiments built around concrete problems rather than technology for its own sake.
Global winner · Microsoft AI Hackathon
Multi-tenant disaster early-warning platform that coordinates seven specialized agents over real-time earthquake, fire, and flood data from eight sources.
Python / TypeScript / Microsoft Agent Framework / Azure OpenAI
1st place · Hack Latam
Satellite geointelligence and risk-monitoring platform combining imagery with hydrological models to anticipate flooding for regional authorities in Peru.
Next.js 16 / FastAPI / PostgreSQL / PostGIS / Google Earth Engine
IEEE research · Machine learning
Production-ready Android application for identifying potential smartphone-addiction cases in school-aged adolescents, backed by a model with 92% accuracy and 0.998 AUC.
Machine learning / Android / IEEE research
Education · Applied AI
Gamified learning platform for primary-school students in Peru, with guided study, text-to-speech, and culturally relevant science content.
React / TypeScript / Supabase / Azure OpenAI / ElevenLabs
1st place · BCP Hackathon
AI-assisted mobile concept designed to help vulnerable regions in Peru prepare for and respond to natural disasters.
AI/ML / Mobile / Python
Developer tool · Browser extension
Chrome extension that asks people to define an intention before entering distracting websites, then keeps that intention visible.
TypeScript / Chrome Extension API / Vite / Manifest V3
Financial software · Web platform
Platform for generating financial lease reports with the German amortization method; I led API design, data modeling, and frontend experience.
JavaScript / C# / .NET Core / Entity Framework